Nuclear accessory factors enhance the binding of progesterone receptor to specific target DNA

~. Sterotd Btochem. Molec. B:ol Vol 48, No. 1, pp. 1-13, 1994

Copyright © 1994 Elsevier Scmnce Ltd Pnnted m Great Bntmn All nghts reserved 0960-0760194 $24,00 + 0.00

Pergamon

0960-0760(93)E0019-4

General Review Nuclear Accessory Factors Enhance the Binding of Progesterone Receptor to Specific Target D N A Paul Prendergast, Sergio A. Ofiate,* Kurt Christensen and Dean P. Edwardst Department of Pathology, University of Colorado Cancer Center, Denver, CO 80262-0216, U.S.A.

The h u m a n p r o g e s t e r o n e r e c e p t o r (PR) is d e p e n d e n t u p o n h o r m o n e a n d a n u c l e a r a c c e s s o r y factor(s) for m a x i m a l b i n d i n g to p r o g e s t e r o n e response e l e m e n t s (PRES) in vitro. R e c o m b i n a n t f u l l - l e n g t h PR, expressed in a b a c u l o v i r u s s y s t e m a n d p u r i f i e d to a p p a r e n t h o m o g e n e i t y , was u s e d as a s u b s t r a t e to isolate a n d i d e n t i f y the accessory factor(s). The m a j o r P R E b i n d i n g e n h a n c e m e n t a c t i v i t y p r e s e n t in n u c l e a r e x t r a c t s was shown to be associated with the high m o b i l i t y g r o u p c h r o m a t i n p r o t e i n HMG-1. Moreover, HMG-1 was e q u a l l y effective in e n h a n c i n g the DNA b i n d i n g o f b o t h the A a n d B i s o f o r m s o f PR. E n h a n c e m e n t of P R E b i n d i n g was highly selective for HMG-1 as a single p u r i f i e d p r o t e i n a n d was not m i m i c k e d by a general p r o t e i n stabilization effect. In gel m o b i l i t y shift assays, it a p p e a r e d t h a t HMG-1 e n h a n c e d P R E b i n d i n g w i t h o u t stably p a r t i c i p a t i n g as a c o m p o n e n t o f the final D N A - P R complex, suggesting t h a t HMG-1 acts i n d i r e c t l y by m o d i f y i n g the P R p r o t e i n or the t a r g e t DNA. HMG-1 is a s e q u e n c e - i n d e p e n d e n t DNA b i n d i n g p r o t e i n t h a t recognizes d i s t o r t e d DNA s t r u c t u r e s a n d is also able to p r o m o t e f u r t h e r d i s t o r t i o n s by b e n d i n g DNA. E n h a n c e m e n t o f P R E b i n d i n g was f o u n d to be intrinsic to the c o n s e r v e d DNA b i n d i n g d o m a i n o f HMG-1 suggesting t h a t HMG-1 acts by p r o m o t i n g a s t r u c t u r a l a l t e r a t i o n in the t a r g e t P R E - D N A .

J. Steroid Biochem. Molec. Biol., Vol. 48, No. 1, pp. 1-13, 1994

INTRODUCTION Steroid hormone receptors are a family of ligand dependent transcriptional activators that function by binding to cis-acting hormone response elements (HREs) of target genes [1]. Major unanswered questions in our understanding of steroid hormone action are how the ligand converts receptors from an inactive to an active transcriptional regulator and how receptor binding to HREs increases transcription of target genes. Although consensus nucleotide sequences for functional HREs and structural features in the receptor that are important for binding have been defined, much remains to be learned about receptor-DNA recognition *Present address' Cell Biology Department, Baylor College of Medicine, Houston, T X 77030, U.S.A. tCorrespondence to D. P. Edwards at the Department of Pathology (B216), University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Denver, CO 80262-0216, U.S.A. Received 18 Oct. 1993; accepted 5 Nov. 1993.

mechanisms. A recurring theme in the literature is the existence of nuclear auxiliary proteins that modulate the in vitro DNA binding activity of sequence specific transcription activators. In some cases, these auxiliary proteins have also been shown to effect transcriptional activity in vivo. Accessory proteins that facilitate DNA binding have been described for the activated T-cell transcription factor N F - A T [2], serum response factor SRF [3],fos-jun [4], NF-kB [5-7], CREB/ATF [7] and several members of the steroid receptor family [8-22]. It appears that there may be common auxiliary proteins for distinct subgroups of transcription activators. For example, the DNA binding activities offos-jun, NFkB, Myb, and CREB/ATF, are sensitive to oxidation-reduction and a protein termed redox factor ref-1 has been cloned and isolated that is capable of mediating both redox regulation and enhancement of the DNA binding of these transcription factors [7]. The retinoic acid X receptor (RXR), which is itself a member of the steroid receptor family, appears to serve

2

Paul Prendergast et al.

in this capacity for a subgroup of intracellular receptors that includes thyroid hormone receptor (TR), vitamin D 3 receptor (VDR), retinoic acid receptor (RAR) and two orphan receptors, the peroxisome proliferator activator receptor (PPAR) and C O U P - T F [12-15, 23, 24]. RXR is a former orphan receptor since a stereoisomer of all-tram retinoic acid, 9-cis retinoic acid, has now been identified as the endogenous ligand for RXR [25]. It appears that RXR enhances DNA binding by heterodimerization where both RXR and the other receptor partner are both components of the D N A complex. Whether proteins other than RXRs can also heterodimerize with and enhance DNA binding of this subgroup of receptors remains to be resolved [8-11]. Nuclear proteins that enhance the DNA binding of the classical steroid hormone receptors have also been described. Two non-histone chromosomal proteins of 60-70 kDa that bind to the spacer nucleotides of palindromic estrogen response elements (EREs) were shown to enhance estrogen receptor (ER) binding [18]. It has also been shown that efficient binding of highly purified human ER to an ERE in vitro requires the addition of a 45 kDa single stranded DNA binding protein [19]. A small heat stable 3-7 kDa polypeptide was reported to increase the binding of glucocorticoid receptor (GR) to DNA-cellulose [16] and a larger protein of 93 kDa was described to increase binding of GR to chromatin [17]. More recently, a receptor accessory factor (RAF) associated with a protein of 130 kDa, was demonstrated to enhance the binding of recombinant GR and androgen receptor (AR) to their cognate target DNA sequences [20]. It appeared that enhancement of DNA binding was associated with RAF forming a ternary complex with receptor and DNA. We have shown in earlier studies that the human progesterone receptor (PR) also requires an auxiliary nuclear protein for maximal binding to progesterone response elements (PREs) in vitro [21]. The major PRE enhancement activity present in nuclear extracts was identified to be associated with the high mobility group chromatin protein HMG-1 [22]. T he present paper provides an overview of our studies of auxiliary proteins that enhance the DNA binding of PR and includes both published [21,22], and unpublished results. A mechanism is proposed by which accessory factors enhance receptor binding to specific DNA sites by promoting structural alterations in the target DNA. EXPERIMENTAL

Cell culture, baculovirus expression and receptor preparation PR-rich T47D human breast cancer cells and PRnegative MDA231 cells were cultured as described previously [26, 27]. S. frugiperda (Sf9) insect cells were grown at 27°C in Grace's insect medium (GIBCO) supplemented with 3.3g/1 of yeastolate (GIBCO), 3.3 g/1 of lactalbumin hydrolysate (GIBCO), 25/~g/ml Gentamicin (Irvine) and 10% fetal bovine serum (Hyclone). For infection with recombinant baculovirus

vectors, Sf9 cells were plated at a density of 1 × 1 0 6 cells/ml in spinner vessels. Cells in suspension culture were infected with recombinant viruses at a multiplicity of infection of 2.0 for 48 h at 27°C. Construction of recombinant baculoviruses expressing either the intact full length A or B isoforms of human PR and their functional properties have been detailed previously [28]. Recombinant receptors expressed from baculovirus were prepared as whole cell extracts (WCE) of infected Sf9 cells by lysing cells in T E D G [10 mM Tris-base, pH 7.4, 1 m M E D T A , I mM dithiothreitol ( D T T ) , 10% glycerol] supplemented with a cocktail of protease inhibitor [26] and 0.5 M NaC1. Samples were clarified by centrifugation at 105,000g for 30 min. Receptor binding to hormone in vivo was accomplished by incubating Sf9 insect cells at 27°C for the last 4 h of infection with 100nM of the synthetic progestin, R5020. WCEs were dialyzed at 4°C against T E D G to reduce salt concentrations prior to immunoprecipitation and DNA binding assays. Receptor concentrations in cell extracts were estimated by steroid binding and by immunodot blot assay as described previously [29], and protein concentrations were measured by the method of Bradford [30]. Purification of recombinant human P R Separately expressed PR-A and PR-B from baculovirus vectors in Sf9 insect cells were purified by immunoaffinity chromatography. Monoclonal antibody (MAb) affinity resins were constructed by chemically crosslinking the B-30 or the AB-52 MAb to protein G Sepharose (Pharmacia-LKB) using 10 mM dimethylpimelimidate by methods described previously [26]. MAbs were purified from ascites fluids and coupled to protein G-Sepharose at a concentration of a 4-6 mg/ml. T he B-30 MAb which recognizes the B isoform of PR was used to purify PR-B. AB-52 MAb which recognizes both the A and B isoforms was used to purify PR-A [26]. Approximately 300 x 106 Sf9 insect cells were used for each receptor purification. T o follow receptor during purification, WCEs were incubated for 8 h at 4°C with [3H]RS020 to allow some exchange with unlabeled R5020 bound to PR. Excess free [3H]RS020 and NaC1 concentrations in WCEs were reduced by dialysis against T E D G . Dialyzed WCEs were then incubated for 4 h at 4°C on an end-over-end rotator with a 1.0ml suspension of MAb-coupled protein G-Sepharose. The beads were then collected by centrifugation at 1500 rpm for 5 min and washed by resuspension and centrifugation in 15 ml of T E G ( T E D G minus D T T ) . This was followed by 3 more washes with 15 ml of T E G containing 0.5 M NaC1 and one additional wash with 15 ml of T E G . Resins were then transferred to a new tube and washed twice more with 15ml of T E G . For elution of PR, resins were transferred to a 2.0 ml mlcrofuge tube and exposed to alkaline pH. Briefly, beads were suspended in 600 # 1 of 50 mM Tris-base, 1 m M E D T A containing 20% glycerol that was adjusted to pH 11.3 with NaOH. The

Progesterone Receptor-DNA Interactions beads were immediately centrifuged and the supernatant was removed and mixed with 200#1 of a neutralization/renaturation buffer comprised of 400 mM Tris-HC1 (pH 7.4), 40 mM MgCI2, 40 mM D T T , 4 mM EDTA, 0.4 m M EGTA, 100 mM NaC1, 0.2 m M ZnC12 and 50°/0 glycerol. Alkaline pH elution steps were repeated 4--6 times and the separate eluates were pooled. The majority of hormone (60°/0) originally bound to PR remains bound during alkaline pH elution and neutralization (data not shown). Purified PR can be frozen at - 7 0 ° C in aliquots and will retain biological activity for approx. 1-2 months, if not repeatedly freeze-thawed.

Gel electrophoresis and Western blotting SDS polyacrylamide gel electrophoresis (SDSPAGE) and Western blot analyses of PR were carried out as described previously [26, 27, 29] except that peroxidase conjugated anti-mouse IgG and ELISA was used as the detection method instead of radioactive protein A and autoradiography. Silver staining of SDS-PAGE was performed by the method of Heukeshoven and Dernick [31].

Purification of H M G - 1 HMG-1 was purified from calf thymus by a slight modification of the method of Adachi et al. [32]. Briefly, acid soluble chromosomal proteins were extracted with 5% perchloric acid and precipitated with cold 255/0 trichloroacetic acid. The precipitate was pelleted by centrifugation at 10,000 rpm for 20rain. The pellet was dissolved in water and dialyzed against 10mM Tris-HCl, pH 7.8. The dialysate was applied directly to a PBE94 polybuffer-exchange resin (Pharmacia-LKB) used under anion exchange conditions. The PBE94 column was washed in 10 mM Tris-HC1, pH 7.8 to remove weakly bound proteins and then eluted with a linear 0-1.5 M NaC1 gradient prepared in 10mM Tris-HCl, pH7.8. Eluted fractions were analyzed by SDS-PAGE and silver staining and by Western blotting with a rabbit anti-sera to HMG-1. Western blot of HMG-1 was performed as for PR except that the detection method was by autoradiography using an 125I-donkey anti-rabbit IgG (11 #Ci/#g, Amersham). The rabbit anti-sera to HMG-1 [33] was diluted 1:150 in Western blot dilution buffer and the 125I-secondary antibody was diluted 1:2000 (~ 100,000 dpm/ml).

Quantitative immunodot blot assay To determine the quantity of PR protein present in WCEs and after purification, an immunodot blot assay was used as described previously with some minor modifications [29]. As a standard for known amounts of PR, cytosols from T47D breast cancer cells were measured for the number of PR sites by steroid binding assay and then serially diluted (2-fold) in T E D G to generate a series of PR concentrations ranging from 2500 to 1.2 fmol/300 #1. An assumption in this assay is

3

that all PR in T47D extracts is capable of binding steroid. Samples of purified PR were initially diluted 1:10 and WCEs were diluted 1:100. Whatman filter paper (3 ram) and nitrocellulose were equilibrated for 30 rain at 25°C (or room temperature) in T E D G and then placed in a Biorad E6-well dot blot manifold. Samples in 300 #1 volumes were applied under vacuum to each well of the manifold and washed once under vacuum with 300#1 TEDG. The nitrocellulose was then removed, blocked with 15/0 bovine serum albumin (BSA)/0.1 5/0 Tween-20 and incubated with 10 #g/ml of AB-52 MAb and processed in the same manner as for Western blot assays. Nitrocellulose filters were incubated with [35S]protein A (Amersham, at 50,000 cpm/ml), dried and exposed to X-ray film. Each spot was quantitated by direct scanning of the radioactivity of the nitrocellulose filter with a Phosphorimager (Molecular Dynamics).

Gel-mobility shift assay DNA-binding assays by gel-mobility shift were performed as described previously with minor modifications [27, 34]. PR (in fmol or ng as indicated in the figure legends) in whole cell extracts or in purified form were incubated for 1 h at 4°C with 0.3 ng of end-labeled [32P]PRE oligonucleotide (sp. act. of 20,000-30,000 cpm/0.1 ng of DNA). The DNA binding reaction also contained poly dA-dT: poly dA-dT as non-specific competitor DNA and 2/~g of gelatin in a DNA binding buffer consisting of 10 mM Tris-base, pH 7.5, 50 mM NaCI, 5 mM D T T , 2 mM MgCI~, and 10°/0 glycerol. The DNA binding reactions (25 pl) were subjected to electrophoresis and autoradiography as described previously [27, 34]. In experiments in which other proteins were added to purified PR, the receptor and protein fractions were mixed and preincubated for 30 min at 4°C prior to addition of DNA. Quantitation of free [~2P]PRE and PRE complexes was carried out by direct scanning of dried gels for radioactivity using the Series 400 Molecular Dynamics Phosphorlmager.

RESULTS

PR requires both hormone and a nuclear accessory protein(s) for efficient binding to specific DATA In human cells, PR is expressed from the same gene as two amino terminal variants termed PR-A (90-94 kDa) and PR-B (118-120 kDa). PR-A is truncated, missing 164 amino terminal residues, and arises from alternative use of a second promoter. Otherwise, PR-A and PR-B have identical sequences in the DNA binding and C-terminal ligand binding domains [35]. Both PR isoforms have been shown to be capable of binding to DNA and to require ligand for induction of binding to specific PREs [34, 36-38]. Additionally, human PR is capable of forming stable dimers in solution and of binding to DNA in three dimeric forms

4

Paul Prendergast et al.

composed of the possible PR subunits; AA, AB and BB dimers [27, 36, 38, 39]. Previous studies from our laboratory have also shown that PR bound and activated by hormone in vivo (intact cells and then extracted from nuclei with 0.45M NaCI), exhibited substantially higher binding activity (5- to 7-fold) for PREs than an equal number of cytosolic receptors bound to hormone and activated by salt treatment in vitro [21, 27, 34, 39]. This suggests that hormone binding modifies PR in a manner in the intact cell that is not mimicked in vitro or, that nuclear factors facilitate PRE binding that are either absent or in low amounts in cytosols. In support of the former possibility, PR is a phosphoprotein and hormone binding in vivo promotes hyperphosphorylation of PR whereas hormone-dependent phosphorylation does not occur in cell-free cytosols [39,40]. However, these are correlative results and there are no direct data as yet to demonstrate a functional role for phosphorylation in receptor-DNA binding. As evidence that the higher DNA binding activity of in vivo activated nuclear PR is due in part to a nuclear factor,

we found that addition of nuclear extracts (lacking PR) to cytosol receptors stimulated a 3- to 4-fold increase in specific PRE binding [21]. The influence of hormone (synthetic progestin R5020) in vitro and in vivo on DNA binding of PR from T47D human breast cancer cells is shown by gel-mobility shift assay in Fig. 1. The target 32P-labeled DNA is an oligonucleotide containing a single PRE from the promoter distal partial palindromic response element of M M T V [1]. Cytosolic PR prepared from T47D cells in the absence of hormone and treated with 0.5 M NaC1 to dissociate receptors from hsp90 failed to bind to the PRE and to form a retarded mobility complex [Fig. I(A), lane 1]. Incubation of salt-treated cytosol PR with R5020 in vitro for 5 h was sufficient to induce PRE binding [Fig. I(A), lane 2]. Incubation with hormone for longer periods of time did not produce a further increase in binding [Fig. l/A), lane 3]. In contrast, PR activated in vivo by the addition of hormone to intact cells and then extracted from nuclei with high salt, displayed a markedly higher level of

ee

A

B

Cyto6~ PR i

CytosolPR (- R5020)

In ritro

R5020

-

PR.I)I~A Complex "

+

+

NE(pg)

+

0

15

30

60

($h) (16h) (Ih)

PR.DNA . Complex

n.s. m

C

cytosol PR (+ R $ 0 2 0 ) NE(pg)

0

15

30

60

PR-DNA . Complex

Fr~

-

Fig. 1. P R b i n d i n g to specific DNA is d e p e n d e n t on h o r m o n e a n d a n accessory n u c l e a r factor. (A) I n vztro a n d in v i v o effect o f h o r m o n e o n b i n d i n g of T47D P R to DNA. U n l i g a n d e d cytosolic P R p r e p a r e d f r o m T47D cells were t r e a t e d w i t h 0.5 M NaCI for 1 h at 4°C a n d i n c u b a t e d w i t h R5020 (200 n M ) a t 4°C for t h e t i m e s i n d i c a t e d . R e c e p t o r s were also b o u n d to R5020 in v i v o (40 riM) for I h at 37°C a n d the PR-R5020 c o m p l e x e s were t h e n e x t r a c t e d f r o m isolated nuclei w i t h 0.5 M NaCI. A n e s t i m a t e d 0.05 p m o l of P R in W C E s (as assessed by h o r m o n e b i n d i n g ) were s u b m i t t e d to gel-shift assay w i t h 0.3 ng of a [32P]PRE oligonucleotide. P R - D N A complexes, non-specific p r o t e i n - D N A c o m p l e x e s a n d free DNA a r e indicated. (B a n d C) Effect of r e c e p t o r - d e p l e t e d n u c l e a r e x t r a c t on t h e D N A b i n d i n g of cytosolic PR. N u c l e a r e x t r a c t s f r o m n o n - h o r m o n e t r e a t e d T47D cells were dialyzed a n d i m m u n o d e p l e t e d of P R b y r e c e p t o r - s p e c i f i c M A b affinity resins. R e c e p t o r d e p l e t i o n was effective as d e m o n s t r a t e d by W e s t e r n i m m u n o b l o t w i t h r e c e p t o r - s p e c i f i c M A b (not shown). D e p l e t e d n u c l e a r e x t r a c t s (in /~g of t o t a l p r o t e i n s ) were m i x e d in a D N A - b i n d i n g buffer w i t h 0.05 p m o l of e i t h e r u n l i g a n d e d (--R5020) cytosolic P R (B) or with l i g a n d e d ( + R 5 0 2 0 ) cytosolic P R (C) a n d i n c u b a t e d for a n o t h e r 30 m i n at 4°C. [32P]PRE was t h e n a d d e d for a n o t h e r 30 r a i n at 4°C a n d t h e e n t i r e r e a c t i o n m i x t u r e (20 pl) s u b m i t t e d to g e l - e l e c t r o p h o r e s i s . B a n d C show only t h e r e g i o n of t h e gel c o n t a i n i n g t h e P R - D N A complex.

Progesterone Receptor-DNA Interactions DNA binding (5- to 10-fold). To determine whether this difference could be due to the presence o L a n accessory nuclear protein mixing experiments were done. Nuclear extracts from non-hormone treated T47D cells were depleted of PR by MAb-alTlnity chromatography as described previously [21]. Aliquots of the PR-depleted nuclear extracts were then mixed with cytosolic PR and tested for effects on PR-DNA binding. Unliganded cytosolic PR continued to show no DNA binding activity in the absence or presence of added nuclear extracts [Fig. I(B)]. In contrast, addition of nuclear extracts to cytosolic PR bound to R5020, stimulated binding to higher levels than that induced by the hormone alone [Fig. 1(C)]. This stimulatory effect was dependent on the amounts of protein added and appeared to reach saturation, indicating that a binding phenomenon was responsible for the enhancement activity. These results suggest that the higher DNA binding activity of PR bound to hormone in vivo is due, at least in part, to an auxiliary nuclear factor. To further characterize the nuclear factor we have purified human PR to apparent homogeneity and have used the purified receptor as a substrate for addition of accessory factors. Because endogenous PR in mammalian cells is expressed at low levels, recombinant receptor overexpressed in a baculovirus insect cell system was used as a source for purification. Both PR isoforms were expressed separately as intact full-length proteins, and they have been shown previously to be structurally and functionally indistinguishable from their endogenous counterparts [28]. The recombinant PRs bind hormone with the same affinity and steroid specificity as endogenous PR and when present in whole cell extracts they also bind with high affinity to PREs in a hormone-dependent manner [28]. Each PR isoform was purified by MAb affinity chromatography from whole cell extracts of baculovirus infected Sf9 insect cells. For simplicity of presentation, most results in this paper are shown with PR-B. Essentially identical results were obtained with PR-A (see Fig. 6). PR was eluted from the MAb affinity resin by exposure to alkaline pH and was immediately neutralized with a buffer containing excess reducing agent, zinc, and glycerol to attempt to renature any unfolded PR. As shown by silver stained SDS-PAGE in Fig. 2 (left panel), WCEs contain high levels of soluble PR-B, detectable as major stainable bands at l lS-120kDa, and the affinity purified receptors are essentially homogenous. The Western blot in Fig. 2 (right panel) confirms the identity of the silver stained bands in WCEs and after purification as authentic PR-B. The band heterogeneity of PR-B on SDS-gels that is evident in Fig. 2 is due to phosphorylation [28]. It should be noted that PR was bound to hormone in vivo prior to cell lysis and MAb-affinity resins were washed in buffers containing 0.5 M NaCI. These are conditions that favor receptor dissociation from heat shock proteins and immunophilins (p59) that associate with the inactive cytosolic receptor complex [41 ]. Thus

5

200-

PR-B

116-

68-

45 -

1

2

3

Silver staining

4

5

Western blot

Fig. 2. Purification of recombinant PR-B by MAb affinity chromatography. WCEs from Sf9 insect cells expressing h u m a n PR-B were prepared and PR was purified by a receptor-specific MAb (B-30) linked to protein G-Sepharose as described in the Experimental section. PR in WCE (25 #g of total protein) and after purification (5 pg) were analyzed by S D S - P A G E and silver staining (left panel) and by Western blot (right panel) adapted from Ref. [22].

purified PR appears to be essentially free of contaminants and other known receptor-associated proteins. PR in WCEs and after purification was quantitated by steroid binding assay and by dot blot immunoassay. Approximately 80% of PR protein in WCEs can be accounted for as having steroid binding activity (data not shown). As found previously with endogenous T47D PR [26], exposure to alkaline pH is a relatively mild denaturation condition that maintains the majority (>60%) of the steroid originally bound to the native receptor (not shown). PR in WCEs and after purification were also analyzed for their ability to bind to specific DNA by gel-mobility shift assay. As shown in Fig. 3 (left panel, lane 2), hormone (R5020) activated PR-B present in WCEs bound efficiently to the PRE oligonucleotide. By contrast purified PR-B exhibited no DNA binding activity under the conditions of the gel shift assay (Fig. 3, left panel, lane 3). It should be noted that equal amounts (estimated by dot blot immunoassay and steroid binding) of PR in WCEs and in purified form were added to DNA-binding reactions to ensure that differences in DNA binding were not due to different amounts of receptor protein. Remarkably, the DNA binding activity of purified PR-B was nearly fully restored to that of native PR by addition of 0.45 M NaC1 nuclear extracts that lacked PR [Fig. 3, left panel,

6

Paul Prendergast et al.

lanes 3-8]. Moreover, the ability to stimulate DNA binding was dependent on the amount (in #g of total protein) of nuclear extract added and appeared to be saturable. The reconstituted complex contained PR-B as demonstrated by a further mobility shift induced by addition of a PR-B specific MAb ([22], not shown). Nuclear extracts prepared from several different eukaryotic cells enhanced the DNA binding of purified PR-B including, PR-positive T47D breast cancer cells (PR was immunodepleted), PR-negative MDA-231 breast cancer cells and non-infected Sf9 insect cells (Fig, 4). Activity was also present in rabbit reticulocyte lysates (not shown), but little or no activity was detectable in E. coli extracts (see Fig. 6). This suggests that the PRE binding enhancement activity is associated with a ubiquitous eukaryotic factor. By subcellular fractionation the majority of activity was present in 0.45 M NaC1 extracts of isolated nuclei, although some activity was found in cytosol (data not shown). Activity also appeared to be associated with a protein(s) since it was inactivated by heating or by trypsin digestion and was retained by dialysis with membranes having molecular weight cut-offs in the range of 12-14,000 (data not shown). To determine whether enhancement of PRE binding was due to a specific protein, as opposed to a general effect of protein concentration on stabilization of highly purified receptor, we examined the ability of various purified proteins and other agents to enhance the DNA

binding of purified PR-B. As shown in Fig. 3, addition of ovalbumin in amounts equal to that of the total protein of nuclear extracts, did not stimulate PRE binding. Table 1 lists other proteins and compounds tested which included charged and neutral polyamino acids, nucleotide triphosphates, excess reducing agents and 5 ~/o polyethylene glycol as a water excluding agent. Of all the agents above, only nuclear extracts were able to effectively stimulate and restore the DNA binding activity of purified PR. These results indicate that a specific protein(s) present in nuclear extracts is responsible for enhancement of PRE binding. Nuclear accessory proteins do not affect steroid binding activity, dimerization, or D N A binding specificity Since some of the steroid may have been stripped from receptors during affinity purification coupled with the fact that ligand is needed for DNA binding, we questioned whether exposure of purified receptor to additional ligand might have a stimulatory effect on the PRE binding. As shown in Fig. 3 (right panel), addition of excess free ligand to purified PR did not stimulate PRE binding, nor did excess ligand influence the stimulatory effect of nuclear extracts. We also questioned whether nuclear extracts might enhance DNA binding indirectly by stimulating the steroid binding activity of purified PR. However, addition of nuclear extracts did not increase the steroid binding capacity of purified PR-B (data not shown).

-- Excess hormone DBSA (pg)

4, Excess hormone

Oval (pg)

DBSA (pg)

~: 0.0 0.4 0.6 0.8 1.0 2.0 0.4 0.8 1.6 2.0

PR-B-DNA-

Oval (pg)

~: 0.0 0.4 0.6 0.8 1.0 2.0 0.4 0.8 1.6 2.0

PR-B-DNA-

Free,

Free1

2

3

4

5

6

7

8

9

10 11 12

1

2

3

4

5

6

7

8

9

10 11 12

Fig. 3. D N A binding o f P R - B in whole cell extracts and after affinity purification. Left panel: equal a m o u n t s (5 n m ) of r e c o m b i n a n t baculovirus e x p r e s s e d P R - B in WCE (lane 2) or affinity purified r e c o m b i n a n t P R - B (lane 3) were a n a l y z e d by gel shift with an e n d - l a b e l e d P R E oligonucleotide as in Fig. 1. In lanes 4-8, increasing a m o u n t s (in p g o f total protein) o f nuclear extracts were added to purified P R - B . Nuclear extracts containing D N A - b i n d i n g s t i m u l a t o r y activity ( D B S A ) were p r e p a r e d f r o m P R - n e g a t i v e MDA-231 breast cancer cells. A non-specific carrier protein, o v a l b u m i n , was added in increasing a m o u n t s to purified P R - B (lanes 9-12). Right panel: the identical s a m p l e s were i n c u b a t e d with excess (200 nM) h o r m o n e (R5020) (adapted f r o m Ref. [22]).

Progesterone Receptor-DNA Interactions T47D (PR-)

MDA

Sf-9

+

+

+

+

-

+

NE

+

+

Purified PR

+

+

PR-B DNA

+

-

n.so

m

n°~

m

Free

+

-

1

2

3

4

$

6

7

8

9

Fig. 4. P R E - b i n d i n g e n h a n c e m e n t activity is present in P R positive and -negative cells. Nuclear extracts were prepared f r o m two breast cancer cell lines, T47D (depleted o f P R by receptor-specific M A b resins) and MDA-231 (PR-negative) and f r o m non-infected Sf9 insect cells. Equal a m o u n t s of r e c o m b i n a n t P R - B in WCEs (lane 2) or affinity purified P R (lane 3) were assayed for binding to a PRE oligonucleotide by gel shift. In lanes 4-9, nuclear extracts f r o m T47D, MDA-231, and Sf9 cells were added to purified P R - B or assayed alone (adapted f r o m Ref. [22] ). Non-specific D N A c o m p l e x e s and free D N A are indicated.

Various steroid hormone receptors have been shown to dimerize in solution in the absence of DNA and the ability to dimerize has been correlated with DNA binding activity [36, 39, 42-44]. To investigate the possibility that the nuclear accessory protein(s) mediates its effect by modulation of PR dimerization, we have employed a coimmunoprecipitation assay described previously that detects dimerization between truncated PR-A and full-length PR-B [36, 39]. The A Table 1. Agents that do not stimulate DNA binding of purified PR General proteins Nucleoudes

BSA Ovalbumin Insuhn RNAse Gelatin Histone H1 Polyammo acids Poly-L-lysme Poly-L-aspartlcacid Poly-L-proline

ATP GTP Reducing agents tiME DTT Water excluding agent PEG

7

and B receptors were expressed separately in Sf9 insect cells and purified by MAb affinity chromatography as in Fig. 2. Varying amounts of purified PR-A and PR-B (in a constant ratio) were mixed together in the presence or absence of nuclear extracts and samples were immunoprecipitated with the B-30 MAb that recognizes an epitope present in the amino terminus unique to PR-B. Coimmunoprecipitation of PR-A with PR-B is indicative of dimerization between the A and B receptors. The amount of PR-A that specifically coimmunoprecipitated with PR-B was the same at all concentrations of purified PR whether or not nuclear extracts were added (not shown, see Ref. [22]). Thus, stimulation of PR-DNA binding activity by the nuclear factor(s) does not appear to occur by modulation of receptor dimerization. Standard gel-mobility shift assays contain excess unlabeled competitor DNA to minimize binding to the specific 32p-labeled oligonucleotide by other nonspecific DNA binding proteins present in crude nuclear extracts. We thought it important to determine whether nuclear extracts might stimulate the binding of purified PR to specific PREs by simply displacing PR from competitor DNA. However, when the amount of competitor DNA in gel shift assays was reduced or eliminated, purified PR still failed to bind to the PRE under the conditions of the assay and continued to require addition of nuclear proteins (data not shown). With purified PR we routinely used 80 ng of poly dAdT: poly dA-dT as competitor DNA which is considerably reduced from the more standard amounts of 1-2pg. The above results collectively are consistent with the presence of a specific protein in nuclear extracts that enhances the DNA binding of PR for its response element DNA. The high mobility group protein H M G - I functionally substitutes for nuclear extracts

Using nuclear extracts prepared from PR-negative MDA-231 breast cancer cells as a source, we attempted to purify a protein associated with PRE binding stimulatory activity. Conventional column chromatography was used and column fractions were assayed for their ability to enhance the binding of purified PR to PREs by gel shift assay. The majority of PRE enhancement activity was retained by heparin Sepharose and phosphocellulose and required high salt concentrations for elution; 0.8 M NaCI for elution from heparin Sepharose and 0.6M for elution from phosphocellulose. Activity bound poorly to double stranded D N A agarose and did not bind at all to a PRE oligonucleotide column. In contrast, the majority of activity bound to denatured single stranded DNA-agarose and was eluted at 0.2 M NaC1 (Fig. 5). Taking advantage of these properties we used sequential column chromatography steps of heparin Sepharose, phosphocellulose and two passes over single-stranded DNA-agarose to obtain an approx. 100-fold enrichment of activity per unit of protein (not shown).

Paul Prendergast et al.

8 ss-DNA Alpu-ose Elution NaCl (M)

PR-B DNA -

n°~.

u

u.$.

i

Fl't~

i

1 2 3 4 5 6 7 8 9 10 II 12 13 Fig. 5. Fractionation of t h e P I l E binding e n h a n c e m e n t activity by single-stranded D N A - a g a r o s e . Nuclear extracts prepared f r o m P R - n e g a t i v e MDA-231 breast cancer cells were sequentially fractionated on heparin Sepharose, phosphocellulose and then applied to a s i n g l e - s t r a n d e d D N A - a g a r o s e c o l u m n . B o u n d m a t e r i a l was eluted stepwise w i t h NaCI at the concentrations indicated. The c o l u m n flow through (FT) and salt eluted fractions (lanes 2-13) were dialyzed and m i x e d w i t h purified r e c o m b i n a n t P R - B in a g e l - m o b i l i t y shift assay (adapted f r o m Ref. [22]).

During the course of these fractionation experiments we noticed that several other known nuclear proteins have been reported to exhibit similar chromatographic behavior, in particular binding to single-stranded DNA-agarose. These proteins include the basal transcription factors T F I I D [45] and T F I I B [46] and the high mobility group chromatin protein HMG-1 [47]. T F I I B was also of interest because of it was recently identified as a protein that stabilizes the DNA binding of an orphan receptor, COUP-TF, and it was shown to bind directly to PR and ER [48]. TBP (TATA binding protein of T F I I D ) and T F I I B were expressed as GST-fusion proteins in E. coli and purified by glutathione-Sepharose affinity chromatography [48]. H M G 1 was purified from calf thymus by ion exchange chromatography on polybuffer (PBE 94) chromatofocusing resins [32]. Each purified protein, compared with BSA, E. coli extracts and nuclear extracts prepared from MDA-231 breast cancer cells, was tested for its ability to enhance PRE-binding of purified PR. As shown in Fig. 6, only HMG-1 and nuclear extracts from MDA-231 cells had a substantial effect on enhancing PR-DNA binding. Figure 7 is a silver stained SDS-PAGE of the purified HMG-1 used in these experiments. This shows a single silver stained band at 28,000 kDa which was confirmed immunologically by Western blot to correspond to HMG-1. HMG-1 was purified from two sources, calf thymus and MDA-231

nuclear extracts. On a molar basis, HMG-1 was effective when added in stoichiometric amounts with PR. It also appeared that HMG-1 enhanced PRE binding without participating stably in the DNA-receptor complex. As evidence of this, HMG-1 alone did not directly bind to the PRE oligonucleotide, an HMG-1 antibody failed to induce a further mobility shift under conditions where a PR-specific MAb induced a supershift, and enhancement of PRE binding occurred without an obvious change in mobility of the DNA-PR complex (not shown, see Ref. [22]). HMG-1 harbors two domains termed " H M G - A and -B boxes," that comprise structure-specific sequenceindependent DNA binding motifs [49, 50]. The A and B boxes are homologous (but not identical), 80 amino acid regions located respectively in the N-terminus and central portions of the protein. H M G boxes are proposed to be a novel DNA binding motif that recognizes angles or bends in DNA [49, 50]. Regions of homology to H M G boxes have been found in several other proteins including the sequence specific transcriptional regulators, lymphoid cell enhancer factor LEF-1 [51] and testes determining factor SRY [52]. HMG-1 box domains expressed and purified from E. coli were tested for their ability to stimulate the DNA binding of purified PR. The A box exhibited weak enhancement of PRE binding, whereas the B box exhibited strong activity nearly equivalent to that of intact HMG-1. In contrast, H M G boxes of LEF-1 and SRY failed to enhance PRE binding of purified PR (not shown, see Ref. [22]). Thus the HMG-B box alone is sufficient for enhancement of PRE binding, but it appears that not all H M G box motifs are equivalent in this respect. H M G - I increases the binding affinity of P R

Because essentially no specific PRE complexes were detected with purified PR alone, it was not possible to estimate a fold enhancement of PRE binding promoted by HMG-1. Protein-DNA interactions that exhibit high rates of dissociation are generally not maintained under standard gel shift conditions, which suggested that the affinity of PR alone may be too low for detection under our conditions and that the effect of HMG-1 is to increase PR affinity to a detectable level. In order to estimate the binding affinity of purified PR in the presence and absence of HMG-1, we performed binding analysis with varying concentrations of PR against a constant 32P-labeled PRE DNA present in a limiting amount. Much higher concentrations of receptor were used here than the single concentration (5 nM) of purified receptor used in previous experiments. Specific PRE-receptor complexes were quantitated by direct analysis (Phosphorimager) of the radioactivity of dried mobility shift gels. As shown in Fig. 8, binding of native PR in WCEs to a synthetic palindromic PRE was dose-responsive and saturable in the nM range. Purified PR alone, over the concentration range that gave saturable binding in WCEs, failed to exhibit any detectable PRE binding. However, some PRE complexes were obtained at much

Progesterone Receptor-DNA Interactions higher concentrations of purified receptor, but the binding did not saturate and the number of complexes obtained at the highest concentration of purified receptor was less than that exhibited at saturation with much lower concentrations of native PR in WCEs. Essentially identical results were obtained with PR-A [Fig. 8(A)] and PR-B [Fig. 8(B)]. Based on the concentration of PR that gave a 50% upshift of the [32P]PRE, we have estimated a kDa of 2 × 10-9M for both PR-A and PR-B in WCE and a 10-fold higher kDa of 2 × 10 -s M for purified PR (the A and B isoforms were not significantly different). Addition of nuclear extracts or HMG-1 to purified PR resulted in a leftward shift of the binding curve that was nearly superimposable on the curve exhibited by native PR in WCEs. This indicates that HMG-1 (and nuclear extracts) increased the binding affinity of PR by approx. 10-fold. It is not shown here, but it is important to note that HMG-1 appeared to stimulate the binding affinity of PR for the partial palindromic element of M M T V by > 10-fold. However, the difference in kDa in the presence and absence of HMG-1 was not precisely measurable because PR binds less well to the imperfect than to the perfect palindromic PRE. We had difficulty in obtaining any binding of purified PR alone to the PRE of M M T V , even at the highest concentrations of receptor.

includes TR, VDR, RAR, PPAR and C O U P - T F [12-15, 23, 24], acting to enhance DNA binding by a heterodimerization mechanism. Nuclear accessory factors, distinct from RXR, have been reported to enhance the DNA binding of receptors for the classical steroid hormones including receptors for progesterone [21,22], estrogen [18, 19], glucocorticoids [16, 17, 20] and androgen [20]. In general much less is known about the identity and mechanism of action of these accessory factors. We have demonstrated that HMG-1 can function in this capacity to enhance the binding of purified human PR to its specific PRE and this has been shown when both PR and HMG-1 were purified to apparent homogeneity. It is unclear at present whether the classical steroid hormone receptors (GR, MR, AR, ER, PR), as a group, utilize a common accessory protein such as HMG-1 or whether different accessory factors are involved for each receptor. Based on the reported differences in the literature in molecular mass, it would appear that there are multiple accessory factors [16-22]. However, from most of these studies, because they have used either crude factor and/or receptor preparations, it is difficult to interpret whether activity is associated with the same, related or different proteins. These accessory factors presumably do not function by a heterodimerization mechanism, since the classical steroid hormone receptors bind efficiently to DNA as homodimers [39, 42-44]. In order to determine whether HMG-1 has the ability to enhance the DNA binding of other steroid hormone receptors will

DISCUSSION T h e RXR appears to be a common coregulator for a subgroup of the steroid receptor family that PR-B

WCE

i

P~oONA

PR-B

0.3ug

9

FR-B

1.5Pg

1.$pg

I

m

WCEPR-B I I O

PR-B ~

1.6~

PR-B

PR-B

WCE I I ° ~ . ~ ~

1.0pg

PR-ONA m

w?,

PR-B

PR-B

1.5Pg

W E'7o PR-B

1.2Mg

PR-DNA

Fig. 6. I-IMG-1 enhanced the binding o f p u r i f i e d P R - B to PREs. P R - B expressed in Sf9 cells and b o u n d to h o r m o n e (R5020) in v i v o was prepared either as a W C E or was affinity purified. E q u a l a m o u n t s (25 ng) o f P R - B in WCE or in purified f o r m were assayed by gel shift assay a g a i n s t 0.3 n g of a n i m p e r f e c t p a l i n d r o m i c P R E oligonucleotide derived f r o m M M T V . P u r i f i e d P R - B was assayed alone ( P R - B ) or after addition of increasing a m o u n t s o f E. c o l i e x t r a c t , BSA, p u r i f i e d TFIID, p u r i f i e d TFHB, nuclear extracts f r o m MDA-231 cells or purified H M G - 1 f r o m c a l f t h y m u s . The a m o u n t s i n / I g o f protein added to purified r e c o m b i n a n t P R - B are indicated in the figure. Each panel shows only the region o f the m o b i l i t y shift gel t h a t contains the D N A - r e c e p t o r c o m p l e x ( a d a p t e d f r o m Ref. [22] ).

10

Paul Prendergast et al. Breast Calf Cancer MW Th)tmus Calls [ I II MDA- I z31

~

~ I

97k ~

2OOk

66k 1

100k I

45k

31k

71k

l

43k

I

~) o< = I

l

1

28k

HMG-1 ~

HMG-1

21k - 18k

14k

ii~:i i

Silver Stain

W e s t e r n Blot

Fig. 7. Purification of high m o b i l i t y group protein HMG-1. HMG-1 was purified f r o m calf t h y m u s and f r o m MDA-231 breast cancer cells by fractionation on a polybuffer-resin (PBE94). The PBE94 c o l u m n was used under anion exchange c h r o m a t o g r a p h y conditions. B o u n d proteins were eluted f r o m the c o l u m n by a linear 0-1.5 M NaCI gradient. HMG-1 that elutes at about 0.6 M NaCI was dialyzed and r e c b r o m a t o g r a p h e d in the s a m e m a n n e r on a second PBE94 column. The m o s t purified fraction was analyzed by 12.5% S D S - P A G E and silver staining and by Western blot w i t h a rabbit polyclonal antibody raised to c a l f t h y m u s HMG-1. lzSI-donkey anti-rabbit IgG secondary antibody and autoradiography was used as the Western blot detection m e t h o d (adapted f r o m Ref. [22]).

require studies similar to those described here, with both purified H M G - 1 and various other purified receptors. Although our studies with H M G - 1 have been conducted in vitro, physiological relevance is suggested by several observations. For instance, enhancement of PRE binding is not confined to recombinant PR and gel-shift as an assay method. As shown here enhancement was observed by gel-shift assay both with endogenous PR from T 4 7 D cells and recombinant PR. In earlier studies, enhancement of P R E binding by nuclear extracts was also detected by an immunoprecipitation assay [21] (immunoprecipitation of P R - D N A complexes with PR-specific MAbs). Purified PR alone is in fact able to bind to a P R E to some extent when receptor ~s added in high enough concentrations (Fig. 8) [22]. T h e effect of H M G - 1 is to increase PR-binding affinity so that r e c e p t o r - D N A interactions occur efficiently at low PR concentrations. Moreover, enhancement of P R E binding did not require excess H M G - 1 , but occurred when H M G - 1 and PR were present in equal molar amounts. Also, H M G - 1 stimulated a greater fold increase of PR binding affinity for an imperfect palindromic PRE, than it did for a synthetic perfect palindromic P R E [22]. In this regard, it is interesting that naturally occurring progesterone

responsive genes contain imperfect palindromic response elements [1]. Although the biological effect of H M G - 1 on PR function in vivo is not known, our data suggest that H M G - 1 could modulate the affinity of receptors for specific D N A sites in vivo and influence transcriptional activity of PR. H M G - 1 is a sequence-independent D N A binding protein that prefers to bind to distorted D N A structures including single-stranded D N A (presumably a stable secondary structure), bends in D N A , cruciform D N A and D N A adducts [33,47,49, 50-53]. Recognition of D N A structure is mediated by a conserved H M G box motif and the H M G box alone is necessary and sufficient for D N A binding. T h e solution structure of the H M G - B box has been solved by 2-D N M R [50]. It is composed of three ~ helices arranged in an L shape with the two arms held at an 80 ° angle to each. This unusual shape has been suggested to be important for recognition of bends or kinks in D N A [50]. Regions of homology to the A and B boxes of H M G - 1 have been discovered in several other proteins including, the sequence specific transcription factors SRY and L E F - 1 [51,52]. Interestingly, the H M G boxes of SRY and L E F - 1 not only brad to distorted D N A structure but also promote sharp bends in DNA. As shown by cyclization assays,

Progesterone Receptor-DNA Interactions HMG-1 also promotes D N A bending but does so in a sequence-independent manner [54]. In fact DNA bending may be a general effect of all proteins with H M G box motifs. T he functional role of H M G - 1 is not well defined. It has been suggested to be involved in replication, DNA repair, recombination and transcription [33, 47, 49-53]. Based on our results with PR, HMG-1 may be a general facilitator of the binding of sequence-specific proteins to DNA. PR-A

100

(A)

WCA-A PR-A+200ng HMG-1 PR-A+500ng NE

80 PR-A m 60

~ 4o .

20

/ e /

0 -- -'~*/* I 0

10

I

I

I

20

30

40

PR Concentration (nM) PR-B

(B)

I00

~•f/

~ 6o 2

e 4020 0

• PR-B+HMO-1 PR-B+500ng NE O WCE-B

•

/* 10

20

30

40

PR Concentration (nM) Fig. 8. H M G - 1 increased the binding affinity of purified P R - A and P R - B for specific D N A . The P R concentration in gels h i f t assays was varied a g a i n s t a constant a m o u n t o f a 32P-synthetic p a l i n d r o m i c P R E oligonucleotide. R e c o m b i nant P R - A and P R - B ( P R - A u p p e r panel, P R - B lower p a n e l ) were assayed in this m a n n e r w h e n p r e p a r e d as a crude WCE or affinity purified. Affinity purified P R was assayed alone ( P R - A or P R - B ) or after addition o f 500 n g o f nuclear extract ( P R - A + NE, P R - B + NE) or 200 ng of purified H M G - 1 ( P R A - / - H M G - 1 , P R - B + H M G - 1 ) . The m o b i l i t y shift gels were directly scanned for radioactivity ( P h o s p h o r i m a g e r , Molecular D y n a m i c s ) and the % upshifted P R E c o m p l e x e s were q u a n t i t a t e d for each concentration o f P R . The m i d p o i n t concentration o f P R r e q u i r e d to bind and upshift 50% o f the D N A was taken as a n a p p r o x i m a t i o n o f the binding dissociation constant (reproduced f r o m Ref. [22]).

11

We propose that high affinity PR binding is dependent upon both nucleotide sequence and protein induced conformational changes in DNA structure. As a working hypothesis receptor alone is predicted to bind with low affinity to PREs to produce a transient distortion in DNA structure. This marks the DNA as a binding site for the sequence-independent structurespecific HMG-1. HMG-1 increases the binding affinity of PR by either stabilizing a dynamic structural conformation in the target PRE that is favored by PR or, by promoting further distortions in DNA structure. HMG-1 interaction appears to be transient or unstable because it is not a component of the final DNA-receptor complex. This model provides an explanation for how a non-specific DNA binding protein like H M G - I can facilitate binding of a sequence specific protein. It will be important in future studies to determine the nature of the structural changes in DNA that are promoted by the combined actions of PR and HMG-1. Recent studies on the ability of steroid receptors to bend DNA provide support for this model. The ER when bound to an ERE was reported to bend DNA, however, this was observed with an E R - D N A binding domain fragment and partially purified full-length ER, but not with highly purified full-length ER that failed to bind to EREs at all. This suggests that an accessory factor may be required both for high affinity DNA interaction and for DNA bending [55, 56]. The thyroid hormone receptor has also been shown to bend DNA when bound to its response element and bending was found to be enhanced by nuclear factors of unknown identity [57]. This raises the possibility that H M G - I or H M G related proteins may act to facilitate DNA bending promoted by steroid receptors and that bending may be requisite for high affinity binding. An important unanswered question is how members of the GR subfamily (GR, MR, AR, PR) can elicit specific responses to their cognate ligands in vivo when each receptor can bind to and enhance transcription from the same consensus glucocorticoid response element (GRE). It has been suggested that cell-specific accessory proteins are involved in specifying responses in vivo on the same GRE [I]. Several proteins with a wide range in molecular mass have been discovered to have H M G box motifs and to bind to DNA structures [33, 47, 49-53]. It is possible that some of the proteins of varying size that have been reported to enhance the DNA binding of the classical steroid receptors may be related to H M G - I by virtue of containing an H M G box motif. In theory other proteins that harbor H M G boxes may be able to functionally substitute for H M G I. In support of this notion is the fact that enhancement of PRE binding was found to be intrinsic to the H M G - B box (see Ref. [22]). Not all H M G boxes, however, appear to be equivalent in this regard since the H M G boxes of L E F - I and SRY failed to enhance P R - D N A binding (see Ref. [22]). Perhaps, differential utilization of H M G related proteins as accessory factors plays a role in dictating specificity within the GR

12

Paul P r e n d e r g a s t et al.

subfamily by promoting alternative structural conformations in the same GRE that will be favored by one receptor. Acknowledgements--The authors are grateful for the generous contributions of Drs Raymond Reeves and Mark Nlssen, Washington State University, Pullman, WA, for providing purified HMG-1 from calf thymus; Dr Jean O. Thomas, Cambridge University, England, for providing H M G box domain peptides; Dr Michael Bustin, National Institutes of Health, Bethesda, M D for rabbit anti-serum to HMG-1; and Drs Ming-Jer Tsm and Bert W O'Malley, Baylor College of Medicine, Houston, T X for recombinant TBP and TFIIB. We also thank Nancy Hart and Clairene Mraz for expert assistance with preparauon of the manuscript and Drs Steven K. Nordeen, David Pettijohn and James Wagner, University of Colorado Health Sciences Center, for insightful discussions during the course of these studies. This work was supported in part by U.S. Public Health Service Grants CA46938, Lucille Markey Charitable Trust, and the University of Colorado Cancer Center Core Grant P30CA46934.

REFERENCES 1 Truss M. and Beato M.: Steroid hormone receptors and interaction with deoxyribonucleic acid and transcription factors. Endocrine Rev. 14 (1993) 459--479. 2. Gramelli-Piperno A and McHugh P.. Characterization of a protein that regulates the DNA binding activity of NF-AT, the nuclear factor of activated T cells. Proc. Nam. Acad. Sci. U.S.A. 88 (1991) 11431-11434. 3 Grueneberg D. A., Natesan S., Alexandre C. and Gilman M Z.: Human and Drosophda homeodomaln proteins that enhance the DNA-binding activity of serum response factor. Science 257 (1992) 1089-1095 4 Abate C., Patel L., Rauscher F. J and Curran T.: Redox regulation of los and jun DNA-bmding activity m vitro Science 249 (1990) 1157-1161. 5. Toledano M. B. and Leonard W. J Modulation of transcription factor NF-kB binding activity by oxidauon-reductlon m wtro Proc. Natn Acad. Sci. U.S.A 88 (1991) 4328-4332. 6. Thanos D. and Maniaus T : The high mobihty group protein H M G I(Y) is required for NF-kB-dependent virus induction of the human IFN-fl gene. Cell 71 (1992) 777-789. 7 Xanthoudakls S , Miao G , Wang F., Pan Y.-C E and Curran T.: Redox activation of los-jun DNA binding activity is mediated by a DNA repmr enzyme. E M B O ffl 11 (1992) 3323-3335. 8. Burnside J , Darling D A and Chin W. W.' A nuclear factor that enhances binding of thyroid hormone receptors to thyroid hormone response elements..y. Bwl. Chem. 265 (1990) 2500-2504. 9. Darling D. S., Beebe J. S., Burnside J., Wmslow E. R. and Chin W W.: 3,5,3'-trilodothyronme (T3) receptor-auxiliary protein (TRAP) binds DNA and forms heterodimers with the T3 receptor. Molec. Endocr. 5 (1991) 73-84. 10. Sone T., Ozono K. and Pike J. W. A 55-kllodalton accessory factor facilitates vitamin D receptor DNA binding. Molec. Endocr. 5 (1991) 1578-1586. 11 Ross T K , Moss V E., Prahl J. M. and DeLuca H F.: A nuclear protein essential for binding of rat 1,25-dflaydroxywtarain D3 receptor to its response elements. Proc Natn Acad. Scz U.S.A. 89 (1992) 256-260 12. Khewer S. A , Umesono K., Mangelsdorf D. J. and Evans R. M.: Retinold X receptor interacts with nuclear receptors an retmoic acid, thyroid hormone and vitamin D3 signalling Nature 355 (1992) 446--449. 13 Leid M., Kastner P. H., Lyons R., Naksham H., Saunders M , Zacharewskt T., Chen J.-Y., Stauh A., Garmer J.-M., Mader S. and Chambon P. Purification, cloning, and RXR identity of the HeLa cell factor with which RAR or T R heterodimerizes to hind target sequences efficiently. Cell 68 (1992) 377-395 14 Yu V C., Delsert C., Andersen B., Holloway J. M., Devary O. V., Naar A. M., Kim S. Y , Boufin J.-M., Glass C. K. and Rosenfeld M. G : RXRB. a coregulator that enhances binding of retmoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements Cell 67 (1991) 1251-1266 15. Zhang X-K., Hoffmann B., Tran P. B-V., Graupner G and Pfahl M Retmoid X receptor is an auxiliary protein for thyroid hormone and retmolc acid receptors. Nature 355 (1992) 441-445.

16. Cavanaugh A. H. and Simons S. S. Jr: Glucocortlcold receptor binding to calf thymus DNA. I. Identification and characterlzation of a macromolecular factor involved in receptor-steroid complex binding to DNA. Biochemistry 29 (1990) 989-996 17 Okamoto K., Isohashi F and Ueda K.: Purification and characterization of an adenosine triphosphate-stimulatory factor that enhances the nuclear binding of activated glucocorticoid-receptor complex from rat liver. EndocrinoloKy 123 (1988) 2752-2761. 18 Feavers I. M., Jmcny J., Moncharmont B., Saluz H. P. and Jost J P. Interaction of two nonhistone proteins with the estradiol response element of the avian vitellogenin gene modulates the binding of estradiol-receptor complex. Proc. Natn. Acad Sci. U.S.A. 84 (1987) 7453-7457 19. Mukher)ee R. and Chambon P.: A single-stranded DNA-bindmg protein promotes the binding of the purified oestrogen receptor to its responsive element. Nucleic Acids Res. 18 (1990) 3713-5716 20 Kupfer S. R., Marschke K B , Wilson E M. and French F. S.: Receptor accessory factor enhances specific DNA binding of androgen and glucocorticold receptors J. Biol. Chem. 268 (1993) 17519-17527. 21. Edwards D, P,, Kuhnel B , Estes P A. and Nordeen S. K.: Human progesterone receptor binding to mouse mammary tumor virus deoxyribonucleic acid: dependence on hormone and nonreceptor nuclear factor(s). Molec. Endocr. 3 (1989) 381-391. 22 Ofiate S. A., Prendergast P , Wagner J P , Nmson M., Reeves R., Pettijohn D. and Edwards D P : Identification of the high mobility group protein HMG-1 as an accessory factor that enhances the specific DNA binding activity of progesterone receptor. Molec. Cell Biol. In press. 23 Khewer S. A., Umesono K., Heyman R. A., Mangelsdorf D. J., Dyck J A and Evans R M.: Retlnold X receptor-COUP-TF interactions modulate retmoic acid signaling. Proe. Natn Acad Sc~ U S.A. 89 (1992) 1448-1452 24 Khewer S. A., Umesono K , Noonan D. J., Heyman R. A and Evans R. M.: Convergence of 9-czs retmoic acid and peromsome proliferator signalling pathways though heterodimer formation of their receptors. Nature 358 (1992) 771-774. 25 Heyman R. A , Mangelsdorf D. J., Dyck J. A., Stein R. B., Elchele G., Evans R. M. and Thaller C.: 9-Cis retmom acid is a high affinity ligand for the retmold X receptor Cell 68 (1992) 397-406. 26. Estes P. A., Suba E. J., Lawler-Hcavner J., Elashry-Stowers D., Wel L. L., Toft D. O., Sullivan W. P., Horwltz K. B. and Edwards D. P.: Immunologic analysis of human breast cancer progesterone receptors. 1 Immunoaffmity purification of transformed receptors and producuon of monoclonal antibodies Bwchemzstry 26 (1987) 6250-6262 27 EI-Ashry D., Ofiate S A , Nordeen S K and Edwards D P.. Human progesterone receptor complexed with the antagonist RU486 binds to hormone response elements m a structurally altered form Molec. Endocr 3 (1989) 1545-1558. 28. Chnstensen K , Estes P A , Ofiate S A , Beck C. A , DeMarzo A , Altmann M., Lleberman B A , St. John J., Nordeen S. K and Edwards D. P.. CharacterJzation and functional properties of the A and B forms of human progesterone receptors synthesized in a baculovlrus system. Molec Endocr. 5 (1991) 1755-1770 29 Elashry-Stowers D., Zava D T., Speers W. C and Edwards D. P.. Immunocytochemical localization of progesterone receptors in breast cancer with anti-human receptor monoclonal antibodies Cancer Res 48 (1988) 6462-6474 30. Bradford M 1976. A rapid and sensitive method for quanutauon of microgram quantities of protein utlhzmg the principles of protein-dye binding. Analyt. Bwchem. 72 (1976) 248-254 31 Heukeshoven J. and Dernick R • A simplified method for silver staining of proteins in polyacrylamide gels and the mechanism of silver staining. Electrophoresis 6 (1985) 103-112 32. Adachi Y , Mlzuno S. and Yoshida M.. Efficient large-scale purification of non-hxstone chromosomal proteins HMG1 and HMG2 by using polybuffer exchanger PBE94. ,y Chromat 530 (1990) 39-46. 33. Emck L. and Bustm M.' 1985. The lntracellular distribution and function of the lugh mobility group chromosomal proteins. Exp Cell Res 156 (1985) 295-310. 34. Ofiate S A , Estes P A., Welch W J., Nordeen S. K. and Edwards D P.. Evidence that hsp-70 associated with progesterone receptors is not involved in receptor-DNA binding Molec Endocr 5 (1991) 1993-2004 35. Kastner P , Krust A., Turcotte B., Stropp U., Tora L , Gronemeyer H and Chambon P Two distract estrogen-regulated

Progesterone Receptor-DNA Interactions

36.

37.

38.

39.

40.

41. 42.

43.

44. 45.

46.

promoters generate transcrips encoding the two functionally different human progesterone receptor forms A and B. E M B O .7/9 (1990) 1603-1614. DeMarzo A. M., Ofiate S. A., Nordeen S. K. and Edwards D. P.: Effects of the steroid antagonist RU486 on dimerization of the human progesterone receptor. Biochennstry 31 (1992) 10491-10501. Bagclu M. K., Tsal S. Y., Tsai M.-J. and O'Malley B. W.: Identification of a functional mtermedmte m receptor actwation in progesterone-dependent cdl-free transcription. Nature 345 (1990) 547-550. Meyer M. E., Parnon A., Bocquel Ji J., Bocquel M., Chambon P. and Gronemeyer H.: Agonistlc and antagonisuc activities of RU486 on the functions of human progesterone receptor. E M B O ffl 12 (1990) 3923-3932. DeMarzo A. M., Beck C. A., Ofiate S. A. and Edwards D. P.: Dmaenzation of mammahan progesterone receptors occurs m the absence of DNA and is related to the release of 90-kDa heat shock protein Proc. Natn. Acad. Scz. U.S.A. 88 (1991) 72-76. Beck C. A., Weigel N. L. and Edwards D P • Effects of hormone and cellular modulators of protein phosphorylation on transcriptional activity, DNA binding, and phosphorylation of human progesterone receptors. Molec Endocr. 6 (1992) 607-620. Smith D. F. and Toft D. O." Steroid receptors and their associated proteins. Molec. Endocr. 7 (1993) 4-11. Fawdl S. E., Lees J. A , White R. and Parker M. G.. Characterization and colocallzataon of steroid binding and dimerezation activities in the mouse estrogen receptor. Cell 60 (1990) 953-962. Rodriguez R., Weigd N. L., O'Malley B. W. and Schrader W. T.: Dimerizataon of the chicken progesterone receptor m vitro can occur in the absence of hormone and DNA Molec. Endocr. 4 (1990) 1782-1790. Kumar V. and Chambon P.: The estrogen receptor binds ughtly to its responsive clement as a ligand-induced homodimer. Cell 55 (1988) 145-156. Nakajima N., Honkoshi M. and Roeder R. G.: Factors revolved m specific transcription by mammalian RNA polymerase II: Purification, genetic specificity, and T A T A box-promoter interactions of T F I I D . Molec. Cell. Bzol. 8 (1988) 4028-4040. Ha I., Lane W. S. and Reinberg D.: Cloning of a human gene encoding the general transcription initiation factor IIB. Nature 352 (1991) 689-695.

13

47. Bustin M., I.elm D. A. and Landsman D.: Structural features of the H M G chromosomal proteins and their genes. Biochim. Biophys. Acta 1049 (1991) 231-243. 48 Ing N. H., Beckman J. M., Tsai S. Y., Tsai M.-J. and O'Malley B. W.: Members of the steroid hormone receptor superfamily interact with T F I I B ($300-II). ft. Bzol. Chem. 267 (1992) 17617-17623. 49. Blanchl M. E., Falclola L., Ferrarl J. and Lilley D. M. J.: The DNA binding site of HMG1 protein is composed of two similar segments ( H M G boxes), both of which have counterparts in other eukaryotlc regulatory proteins. E M B O ffl 11 (1992) 1055-1063. 50. Weir H. M., Kraulis P. J , Hill C. S., Rame A. R. C., Lane E. D. and Thomas J. O.: The structure of the H M G box motif in the B-dommn of HMG1. E M B O ffl 12 (1993) 1311-1319. 51. Glese K., Cox J. and Grosschedl R.: The H M G domain of lymphoid enhancer factor 1 bends DNA and facilitates assembly of functional nucleoprotein structures. Cell 69 (1992) 185-195. 52. Ferrari S., Harley V R., Ponuggia A., Goodfdlow P. N., Lowell-Badge R. and Bianchl M. E.: SRY, lflce HMG-1, recognizes sharp angles in DNA. E M B O ffl 11 (1992) 4497--4506. 53 Bruhn S. L., Pil P. M., Esslgnmarm J. M., Housman D. E. and Llppard S. J.: Isolation and characterization of human eDNA clones encochng a high mobility group box protein that recognizes structural distortions to DNA caused by binding of the anucancer agent cisplatm. Proc. Natn. Acad Sci. U.S.A. 89 (1992) 2307-2311. 54. Paull T. T , HayKinson M. J. and Johnson R. C.: The nonspecific DNA-binding and -bending proteins HMG-1 and H M G - 2 promote the assembly of complex nucleoprotein structures Genes Dev. 7 (1993) 1521-1534. 55. Nardmll A. M., Greene G. L. and Shapiro D. J.' Human estrogen receptor bound to an estrogen response element bends DNA. Molec. Endocr 7 (1993) 331-340. 56. Nardulli A. M. and Shapiro D. J.. Binding of the estrogen receptor DNA-bmding dom~un to the estrogen response dement induces DNA bending. Molec. Cell. Biol. 12 (1992) 2037-2042. 57 King I. N., deSoyza T., Catanzaro D. F. and Lavm T. N ' Thyroid hormone receptor-induced bending of specific DNA sequences is modified by an accessory factor, ft. Biol. Chem. 268 (1992) 495-501.